A crypto engineer has launched a website that lets AI agents hire humans to perform in-person, real-world tasks.
Alex (X handle @AlexanderTw33ts), an engineer who works with UMA Protocol and Across Protocol, demonstrated the site rentahuman.ai in a video posted Monday. The platform allows people to list themselves with an hourly rate and be booked by autonomous AI agents for a range of errands and IRL assignments — from running simple errands and making purchases to attending meetings, taking photos, or signing documents.
Profiles on the site include an eclectic mix of participants, the developer said, naming an OnlyFans creator and an AI startup CEO among those offering services. The homepage bills itself as “the meatspace layer for AI,” using language such as “robots need your body” to explain that physical actions are still beyond purely virtual agents.
Alex said hiring an on-site human from an AI agent can be performed through “one MCP call.” The platform displays available people, a “become rentable” sign-up option, and a growth metric; the site claims nearly 26,000 sign-ups so far, though Alex acknowledged that figure may include duplicate accounts or impersonations and said the team is working to clean the roster.
In an interview on Across Protocol’s Crosschain podcast, Alex emphasized there will be no native cryptocurrency token for the project. “There’s no token; I’m just not into that,” he said, adding that introducing a token could create unnecessary financial risk for users.
Alex also described how he built the site: using what he calls “vibe coding” with a group of Claude-based AI coding agents. He used a technique known as a Ralph loop, which runs coding agents in repeated cycles until they finish a task. According to him, these agent loops can produce deployable code with little human supervision, and his custom Ralph loop was instrumental in creating rentahuman.ai.
The project arrives amid a spate of experimental AI-agent platforms in 2026. One example is Moltbook, a Reddit-style social site designed for AI bots that has drawn attention for odd, bot-generated conversations and emergent behaviors.
The launch raises questions about verification, safety, and ethics — especially when autonomous systems contract people for physical tasks. The developer says he is addressing duplicate and fraudulent accounts and has intentionally avoided tokenizing the platform to limit financial exposure for participants.